Cholera is a contagious disease widely distributed in the world, in particular in the Third World, where, in certain areas, it is endemic. The serious disorders which develop in the intestinal system prove fatal in a high percentage of the recorded cases of the disease.
The etiological agent of cholera is the Gram-negative microorganism Vibrio cholerae (V. cholerae). This colonises the intestinal tract of individuals who have come into contact with it through ingestion of contaminated food or water, and multiplies to very high concentrations. The principal symptom is severe diarrhoea as a result of which the patient can lose as much as 10-15 litres of liquids per day via the faeces. As a result of the severe dehydration and loss of electrolytes, the patient does not withstand the infection in 50-60% of cases, and dies. The diarrhoea caused by V. cholerae is due to the secretion of cholera toxin, CT, which acts by stimulating the activity of the adenylate cyclase enzyme so as to induce disturbances at cell level.
Although cholera can be effectively cured by controlled and intense rehydration, the distribution of a vaccine in desirable with a view to complete control and future eradication of the disease.
At the present time, there exists a vaccination against cholera, consisting of parenteral administration of killed bacteria. Although some countries insist on vaccination against the disease, there are serious doubts as to its real usefulness, given that the current cellular vaccine protects against the consequences of the infection in only 50% of the cases and that the protection is also extremely limited in duration, to less than 6 months.
In Bangladesh, an experimental trial is in progress (1990-92) of an oral vaccine consisting of killed bacteria with the addition of subunit B of cholera toxin, which is known to be highly immunogenic. This product succeeds in inducing lasting protection, without special side effects (Holmgren J., Clemens J., Sack D A., Sanchez J. and Svennerholm A M; “Oral Immunization against cholera” Curr. Top. Microbiol. Immunol. (1988), 146, 197-204).
Cholera toxin resembles the heat labile toxins of enterotoxigenic strains of Escherichia coli in amino acid sequence, structure and mode of action.
The consequences of infection with an enterotoxigenic strain of E. coli are similar to, though less serious than, those of cholera, and consist of severe diarrhoea and intestinal disorders.
The CT and LT toxins all comprise a single A subunit (or protomer A) responsible for the enzymic activity of the toxin (herein CT-A or LT-A) and five identical B subunits (or protomer B) which are involved in the binding of the toxin to the intestinal epithelial cells (herein CT-B or LT-B).
The A subunit penetrates the cell membrane and causes activation of adenylate cyclase by NAD-dependent ADP-ribosylation of a GTP-binding protein which controls the activity of the enzyme. The clinical effect of this is to cause massive fluid loss into the intestine.
Considerable research has been conducted on cholera toxin and the E. coli heat labile toxins.
The sequence of CT is known and has been described (Mekalanos J. J. et al. Nature 306, page 551 (1983)).
The sequence of LT from enterotoxigenic strains of E. coli is, as mentioned, 80% homologous to CT and it too is known and described in the scientific literature. Spicer E. K. et al (Biol. Chem. 257 p. 5716-5721 (1982)) describe the amino acid sequence of the A sub unit of the heat labile toxin from an enterotoxigenic strain of E. coli found in pigs.
A bacterial chromosomal form of LT has been identified and sequenced by Pickett C. L. et al. (J. Bacteriol. 169, 5180-5187, (1987).
The sequence of the A subunit of LT from a strain of E. coli known to affect humans has also been sequenced (Yamamoto et al, J. Biol. Chem., 259, 5037-5044, (1984)).
In view of the potential clinical significance of a vaccine against cholera and enterotoxigenic bacteria there is a continuing and great interest in producing a detoxified toxin capable of immunising against cholera and enterotoxigenic bacteria. The techniques of genetic engineering allow specific mutations to be introduced into the genes encoding the toxins id the production of the mutated toxins using now conventional techniques of gene expression and protein purification.
Various groups have attempted to identify mutations of the genes, which involve loss of the toxicity characteristics of the encoded proteins. The studies are predominantly being carried out in respect of the gene for the toxin LT, from E. coli. 
Harford, S. et al. (Eur. J. Biochem. 183, page 311 (1989)) describe the production of a toxoid by in vitro mutagenesis of the LT-A gene from E. coli pathogenic for pigs. The resulting successful mutation contained a Ser-61-Phe substitution and a Gly-79-Lys substitution, the former being considered the more important. Harford et al. suggest that, because of the similarities between the LT-A genes in E. coli pathogenic to humans and pigs and the CT-A gene, and because the toxins are thought to operate by a common mechanism, it may be possible to produce a cholera holotoxoid by introducing the Ser-61-Phe mutation into the CT-A gene.
Tsuji, T. et al. (J. Biol. Chem. 265, p. 22520 (1990)) describe the mutation of the LT-A gene from plasmid EWD299 to produce a single substitution Glu-112-Lys which affects the toxicity of the mutant LT yet does not change the immunogenicity of the protein.
Grant, C. C. R. et al. (Abstract B289 of the 92nd General Meeting of the American Society for Microbiology, 26-30th May 1992) describe conservative substitutions of histidines at 44 and 70 and tryptophan at 127 in LT-A which result in significant reductions in enzymic activity.
Some work has been conducted on mutations to CT.
Kaslow, H. R. et al. (Abstract B291 of the 92nd General Meeting of the American Society for Microbiology, 26-30th May 1992) describe mutating Asp-9 and His-44 and truncating after amino acid 180 in CT-A which all essentially eliminate activity. Mutating Arg-9 is said to markedly attenuate activity. Mutating other amino acid sites had little effect on toxicity.
Burnette, W. N. et al. (Inf. and Immun. 59(11), 4266-4270, (1991)) describe site-specific mutagenesis of CT-A to produce an Arg-7-Lys mutation paralleling that of a known detoxifying mutation in the A subunit of the Bordetella pertussis toxin. The mutation resulted in the complete abolition of detectable ADP-ribosyltransferase activity.
International patent application WO 92/19265 (Burnette, Kaslow and Amgen Inc.) describes mutations of CT-A at Arg-7, Asp-9, Arg-11, His-44, His-70 and Glu-112.
Mutations at Glu-110 (LT and CT) and Arg-146 (LT) have also been described in the literature (Lobet, Inf. Immun., 2870, 1991; Lai, Biochem. Biophys. Res. Comm. 341 1983; Okamoto J. Bacteriol. 2208, 1988).
The crystal structure of LT has been determined by Sixma et al (Nature, 351, 371-377, May 1991) and confirms the mutatagenesis results described earlier in the literature, explaining structurally the significance of Glu-112 and Ser-61 in activity of the A sub unit and suggesting that His-244, Ser-114 and Arg-54 which are in the immediate neighbourhood may be important for catalysis or recognition.
It is known that the development of toxicity of the A subunits of CT and LT requires proteolytic cleavage of A1 and A2 subunits at around amino acid Arg-192 (Grant et al Inf. & Immun. (1994) 62(10) 4270-4278).
Immunogenic detoxified proteins comprising the amino acid sequence of subunit A of a cholera toxin (CT-A) or a fragment thereof or subunit A of an Escherichia coli heat labile toxin (LT-A) or a fragment thereof, wherein one or more amino acids at, or in positions corresponding to Val-53, Ser-63, Val-97, Tyr-104 or Pro-106 are replaced with another amino acid are disclosed in WO 93/13202 (Biocine Sclavo SpA). optionally the amino acid sequence may include other mutations such as, for example, substitutions at one or more of Arg-7, Asp-9, Arg-11, His-44, Arg-54, Ser-61, His-70, His-107, Glu-110, Glu-112, Ser-114, Trp-127, Arg-146 or Arg-192.
Detoxified mutants of pertussis toxin have been reported to be useful both for direct intranasal vaccination and as a m mucosal adjuvant for other vaccines (Roberts et al. Inf. & Immun. (1995) 63(6) 2100-2108). Published International patent application WO 95/17211 (Biocine SpA) describes the use of detoxified mutants of CT and LT as mucosal adjuvants.